The Epistemology of Depression Part 5 – The Case of Depression under the Statistical Gaze

Hello, and welcome to Against Utopia, a newsletter investigating authoritarian utopianism in science, technology, politics, culture, and medicine, and anti-authoritarian alternatives. This is Issue Five, published October 17th, 2018.

This newsletter now has a Patreon. If you are enjoying it and want to support my work, please do. I will never charge for this newsletter, or my “IP”, regardless.

New subscribers have asked where they can read previous newsletters. They can be found here.

So far, we’ve examined, in a general sense, how medical facts come to be. Despite what we may have been taught about the mythology of the scientific method in school, we’ve seen precisely how much historical and social baggage the practice of medicine has, and that this baggage lingers well into modernity.

Recall that we’ve seen how 19th and early 20th century doctors viewed genealogy as the way to practice medicine by classifying disease. They believed that if they could classify observations scientifically, then they could practice medicine correctly. We then saw how their clinical experience modified classification, such that a gap came to exist between classifications derived from examining the dead, and the observations of the physiology of the living. At almost exactly the right time, the ascendance of statistical methodologies allowed them to bridge this gap via application of the risk factor. The risk factor enabled the emergent structures of medicine to define everyone as having some percentage of illness, which further allowed the pharmaceutical industry to define the ill and the healthy in a statistical sense using the surrogate endpoint, resulting in drugs engineered to target statistical states transitions, with the implicit promise that this maps to physiology.

Taking this all together, we have (1) a way to label the ill and the not ill, (2) a way to justify these labels with scientific methodologies, (3) a way to expand definitions to apply to as many people as possible objectively and hopefully, defensibly, (4) a way to scientifically and mathematically invalidate lived experience, and (5) a methodology to develop drugs for a defined, expanded, and scientifically valid market, that simultaneously is now the principal way that we derive knowledge of physiology, yet can only derive faulty knowledge thereof.

It’s how we end up with serotonin as a “downer” for bears and all mammals it’s ever been studied in, except in humans, where it is the happy molecule, an “upper” (cf. Issue Two).

We will now examine the case of mental disorders in the 19th and 20th century to see how this generalized process applies to depression. Let’s get into it.

Objective Categorization of Mental Illness – Kraepelinian Nosology

In late 19th to early 20th century, psychiatry as a practice started to become formalized, from its quack origins and pet theories, to medical practice and standardization. There were still plenty of doctors (and charlatans) injecting their patients with insulin on a whim, willfully ignoring informed consent, lobotomizing patients, and electrocuting patients with no theory in mind, but a subset of people with MDs began to try and apply the scientific method to what they were observing in asylums and sanitariums.

Emil Kraepelin was one of these MDs. An influential German psychiatrist at the turn of the 20th century, he believed that psychiatric illnesses were principally biological and genetic in nature. He was also a eugenicist, and eugenics at the time was the basic theory of the genetics of disease. Kraepelin endeavored to make the practice of medicine in mental illness as a whole, and depression specifically, more scientific by attempting to apply categories and objective criteria, to theoretical unobserved and undetermined biological states that were assumed to exist, so that all physicians would use the same language to detect the same diagnostic states and hence make the practice of psychiatric medicine “more scientific”. It is important to step back and realize that at this time, the notions of randomization, clinical trials, and selection bias were acknowledged, but were in their infancy. If you saw nothing but women and the maligned classes of Blacks, Latinos, and other non-whites, you were more wont to say that mental illness is more prevalent in these populations, and to diagnose their illnesses as essential to their personhood, and exclude them from polite society. And that’s exactly what he believed – that there was a fixed amount of “illness” in the population, and that a physician’s job was not to heal or change that state, because nothing could be done about it. A physician’s job was to scientifically identify and exclude. That’s part and parcel of what Kraepelin sought to standardize, with science.

To Kraepelin, eugenics was not the problem; the burgeoning practices of psychiatry worldwide were in a diagnostic crisis. It was a common occurrence for psychiatrists to present at conferences and find that what one group was presenting as depression in a clinical study in the UK, was another US group’s definition of schizophrenia, and vice-versa. This fluidity in diagnosis and standards allowed neurologists of the time, who were just glorified spin doctors with the unquestioned authority to shock, lobotomize, and poison people to cure their mental illnesses, to run amok.

This diagnostic crisis only got worse into the 20th century – there were many attempts at controlled experiments in diagnosis, with the American Psychological Association e.g. sending out videos of cases to 20 psychiatrists and psychologists, and receiving back 14 conflicting diagnoses. The most famous case of this is perhaps the Rosenhan Experiment, where a group of practicing psychologists and students infiltrated mental health institutions by faking schizophrenic symptoms that resolved, but were nonetheless forced to admit that they have a mental disorder, and given drugs. But I digress.

It was clear that psychology and psychiatry had a nosological crisis (and continue to have them!), and Kraepelin’s attempt to make psychiatry “clinical” took place under this backdrop, so it was not entirely without merit.

After about a decade and a half of working with the mentally ill, in Kraepelin’s estimation, nothing could be done about depression. He was prolific in this view, and published guide after guide on how to diagnose effectively, and spent much of his time propagandizing his approaches so that doctors and psychiatrists would begin to use the standards and converge on the same diagnoses. Today, Kraepelin is not well known outside of psychiatry, but within the profession, to this day, he is known for being the first real scientific manager of epidemiological mental health, and for setting the standard for how to manage, observe, and record thousands of observations rigorously. His approach would, after being somewhat discredited in the early to mid 20th century, resurface in 1970 in the form of the Diagnostic and Statistical Manual (DSM) of Mental Disorders, and he is recognized for being the major influence behind modern mental health instruments like the HAM-D (Hamilton Depression Rating Scale).

In connecting him to the movements that produced scientific medicine and the pharmaceutical industry, it’s important to note that Kraepelin was perhaps, in my view, the originator of the idea that you do not need to know the underlying biology in order to classify mental disorders, which I contend is also the major epistemological pillar behind the production of scientific knowledge in the pharmaceutical-industrial complex. If you can identify the steps of progression universally, which Kraepelin called “patterns of symptoms”, then you can group these together as “syndromes”, and then psychiatric medicine shifts from merely “symptomatic” to “clinical”. This approach might seem circular in its reasoning, and that’s because it actually is. Kraepelin’s belief boiled down to the pattern of symptoms = disease, and the disease = its pattern of symptoms. These two ideas, that one does not need to know the underlying biology in order to classify illness, and that the symptoms = disease, and the disease = symptoms, laid the groundwork for the pharmaceutical industry to later come along and build a pipeline that could identify “surrogate endpoints” of, in our previous example in Issue Four, Alzheimer’s Disease, without having to understand what actually causes progression in the disease. This was the early form of the surrogate endpoint thinking we discussed in part 4, at work.

Recall that a surrogate endpoint, typically “staging” of a disease, is fashioned in order for the drug to have a target state to work on.  It is assumed that e.g. stage 2 Alzheimer’s is an object that exists in the world that we can identify with a clinical instrument (i.e. a quiz). If we can identify stage 2, and a theoretical (heh) stage 3, then we can fashion a drug whose goal it is to prevent progression from stage 2 to stage 3. Then we can quantify its effect size in a population of clinical trial participants and decide if the drug works.

If we can’t, for example, cure Alzheimer’s because we don’t really understand what causes it, then the theory is that we can just try to prevent progression to a more severe form of the illness by identifying the transition with a standardized dementia questionnaire. We don’t have to understand any of the how – we can just focus on the what. The only unquestioned assumption here is that the progression is real, that it exists in the physiology of the person: if it doesn’t, then the whole theory falls apart.

Now clearly, there are weaknesses to this case that don’t take an entire century to resolve. If all it took to canonize Kraepelin’s approach was the existence of categories, then we could zoom ahead to petrochemicals, dyes, and the pharmaceutical industry’s production of drugs. However, plenty of Kraepelin’s contemporaries called out the tautology of symptoms as disease, the problem of observing classes of disease without physicalizing them in material biology, and other such problems.

Furthermore, Kraepelin’s theories were developed in the background of the burgeoning psychoanalysis movement led by one Sigmund Freud, which was seemingly demonstrating much success explaining the plight of the average person under modernity. There was a bit of a battle brewing, and in order for us to understand how patient experience and context could be ignored in favor of the objective classifications of the psychiatrists, it is important to understand another seminal figure in 20th century psychiatry, Adolf Meyer, and the work he did to popularize and integrate Freudian psychoanalysis with diagnosis, and occupational therapy.

The Reaction to Classification – Adolf Meyer, Patient Experience and Psychoanalysis

When Adolf Meyer received his MD in 1892, Emil Kraepelin’s ideas and scientific practice of psychiatry had just started to take hold in Europe, and Meyer was enthusiastic about applying them in the clinic to real patients. He took up a postdoctoral position at the University of Zurich, and began practicing on patients, putting Kraepelin’s designations in play, learning, iterating, and working at the bleeding edge of psychiatric science. Around the same time that he began to doubt Kraepelin’s classification schemes, he failed to secure an appointment at the University of Zurich, and he instead moved to the United States.

Meyer’s work in the US began to focus on the burgeoning contextual revolution in psychiatry, psychoanalysis, and the work of Sigmund Freud. Meyer was thoroughly Kraepelian in his approach, like most European doctor of the time. In letters to Kraepelin and his personal writings, he (in my opinion, correctly) identified that Kraepelin’s approach was a tautology, in that the symptoms = disease, and the disease = symptoms. How the patient gets the symptoms, how they manifest, their particular form of suffering gives meaning to the disease state in Meyer’s practice, and after he observed Kraepelin at work and mentored under him, he realized that there were multiple holes in his categorical frameworks which he thought could be patched by the inclusion of patient meanings and context.

The principal form of inclusion of patient meaning and context at the time was psychoanalysis. I won’t digress to explain Freud and psychoanalysis, but suffice it to say that the Freudians emphasized the formative influence of early child rearing on adults, of sexuality, and of the interaction between people and their environments. If the eugenicists stressed objectivity and classification, the Freudians agreed to an extent, but stressed that environments generated the maladies that we are classifying, and the maladies themselves were not determined merely by genes or biology. In a way, theirs was a reaction to the very reactionary position that we are born with a certain amount of diseased genes, and that there can be nothing done about this.

In 1908, Meyer started a psychiatric clinic at Johns Hopkins Hospital to begin to implement new ideas, both of Kraepelin and Freud, in the clinic. He decided not to use the clinical model of Emil Kraepelin for classifying disease, but he incorporated some of his practices of observation, standardized record-keeping, and monitoring of pre and post-symptomatic phases of the disorder, in order to build histories and make inferences about likely disease progression. On the other end, Meyer incorporated social and biological factors that affected the personality into the clinical practices under his supervision. One of his primary theses was that modernity was causing a generalized, societal anxiety called neurasthenia, which was actually better understood as a failure to adapt to the demands of modern life. He also thought that instead of specifiable, objective natural disease classes, people mostly suffered as a result of “psychobiological” life situations and he sought to frame mental disorders as reactions to these situations (“A failure to adapt” vs. “you are broken forever”). He was also an early supporter of occupational therapy as a way to help people cope with their life situations, and find meaning in themselves and their communities by productively participating in them.

With his scientific approach to psychoanalysis and classification, and his ability to bring back the patient’s meaning into the context of medical psychiatry, he began to popularize an idea among the burgeoning profession of psychiatry: that everyday people could experience generalized social anxiety and distress, without being insane. This was a HUGE boon that cannot be understated – Meyer, while working to make classifications more scientific, and to create possibilities and de-pathologize mental disorder, expanded the scope of treatable illness from just the “insane”, to everyone.

While helping popularize occupational therapy as a practice was undoubtedly useful and ameliorative for patients, Meyer unknowingly backdoored the expansion of the label of mental illness by allowing any everyday person to come under the influence of the psychiatric profession. He, in effect, democratized psychiatry and its approach to mental illness, lowering the bar to entry to the psychiatrist’s office from the “criminally insane” to the everyday nervous or anxious person.

If you’re following along at home we now have some of the factors needed for a growth market to explode into existence:

  1. An “objective” criterion, with categorization, used to identify the depressed (Kraepelinian nosology)
  2. A way to make the categorization and definition of treatable extend to as many people as possible

​The diagnostic problems still lingered well into Meyer’s time, and it’s hard to engineer drugs for something that even most experts don’t agree is a treatable condition, or the right treatable condition. We needed a way, once and for all, to say who is depressed, and what the criteria are, so that we could make drugs that fit the bill. Perhaps a manual, from a set of experts, blessed by the FDA and the American Psychiatric Association, could finally put this issue to bed…

Integrating Clinical Objectivity with Patient Experience: The Menninger Brothers, Robert Spitzer, and the DSM

Adolf Meyer retired in 1941, just in time to miss the generation-defining opportunity to diagnose and treat people affected by the worst war the world had ever seen. The Menninger brothers, on the other hand, were at the right place at the right time. Karl and his brother William noticed that soldiers who had gone to war and come back often came back with profound psychological trauma. Karl observed, “We must attempt to explain how the observed maladjustment came about and what the meaning of this sudden eccentricity or desperate or aggressive outburst is.” The answer was not found in a classification scheme, it was in “what was behind the symptom,” as Karl put it.

The Menninger brothers, knowingly or unknowingly, agreed with the direction that Adolf Meyer was taking psychiatry in. They believed that factors in the environment influenced the patient’s ability to adapt and cope with stress, or not. This point of view of the organism and its ability to adapt was canonized in the first version of the Diagnostic and Statistical Manual of Mental Disorders, or DSM, which was released in 1952, and featured mental illnesses named as reactions. Everything, from depression (depressive reaction), schizophrenia (schizophrenic reaction), and anxiety (anxiety reaction), was seen as a reaction to a modern environment within which patients were failing to live.

This psychobiological view, as coined by Meyer and expanded by the Menningers, created a conflict between the “scientific” practitioners of observation and classification led by Kraepelin, and the messier therapy-focused practitioners, like the Menningers and Adolf Meyer. If it was indeed factors in the environment that were to blame for the inability to cope exhibited by patients of mental disorders, then, as we indicated in the previous section, everyone can be insane. Instead of using as an opportunity to de-pathologize mental illness, the APA and other governing bodies saw this expansiveness as a conflict to be resolved, because it made diagnosis, identification, and treatment very complicated, and worst of all, it was the source of the murkiness in diagnosis that we identified previously.

It was this key point in time, where psychoanalysts had shown us that patient meaning and individual context were huge inputs into mental disorders, while scientific observation could simultaneously be a huge boon in diagnosis. Instead of using the two approaches to find a middle ground that could present a more holistic view of patient mental health and brain physiology, the authorities took a utopian view, that only one of these is right, only one of these approaches can be scientific, and it is the one that makes health legible, objective, and most importantly, applicable at scale to the entire population.

In 1963, Karl Menninger wrote “Instead of putting so much emphasis on different kinds … of illness, we propose to think of all forms of mental illness as being essentially the same in quality and different quantitatively”, which neatly captures the idea in context, at a time when it was thought that everyone can have mental disorders, it’s just about how much.

Right when this thinking was at its apogee, when psychiatry was facing this crisis in diagnosis seemingly brought about by including too much context, Robert Spitzer happened on the scene, and the first few editions of the DSM settled the fight between classification and patient context.

The first and second editions featured this inclusion of context by framing everything as a reaction to the environment, with the larger understanding that it was simply a failure to adapt to the demands of modern life that caused illness. In due time, it was implied that we would figure out what these shortcomings were, both in the environment and in our own physiology, but the connection between the two was assumed.

Spitzer worked on precisely this problem by refining the nosology of psychiatric medicine. By the time the DSM 3 was up for approval, he and his colleagues had collected more than a decade of clinical signs and symptoms that had been selected precisely for their observability and standardization. Anything that belied a hidden or implied meaning to mental suffering, in effect, anything that brought with it a theory of mind, was excluded. Only what the doctor observed, via senses and discourse, mattered. Spitzer and his colleagues held the position that reaction implies that there was a healthy way of transacting with one’s environment, and that psychiatrists also knew what it was. This, as we’ve seen, was prone to appearing unscientific. In many ways it was, but that was not a bug of the approach, it was a feature.

As we’ve seen through the duration of these newsletter thus far, pretending that we understand a phenomenon by cleaving away the messier parts of it to make it addressable by science is typically done by those with power or authority. It’s also a utopian view, because it seeks to start from a reduced understanding that can be perfected to address all maladies. This is precisely the business that mid to late 20th century psychiatry was involved in – a utopian view of what mental illness looked like, because the real messiness of it would belie the description and prescription of experts. Psychiatrists cut prescription entirely out of the game by saying they did not know how patients should react to their environments, and therefore they were no longer accountable for this. What they were claiming authority over was applying clinical criteria to observable signs of mental disorder, with reliability. By excluding normative claims to mental health buttressed by reaction and psychoanalysis, they were able to exclude all illegible, unscientific practices from psychiatry, so it began to “look” more scientific. This signed over a tremendous amount of power to psychiatrists, eliminated the unreliability of diagnosis (90% of doctors now converged on the same diagnosis, curtailing the effect of the Rosenhan Experiment), and birthed the opportunity for clear criterion, with surrogate endpoints, drugs to address them, of which the market was everyone, because, as we’ve seen, psychoanalysis was shuttered, but the rhetoric with which it expanded the market of pathological mental disorder was not.

Iproniazid, Imipramine, Reserpine and Prozac: How a Shaky Theory Gained Acceptance with Clinical Trials and the FDA

In 1952, doctors in a clinical trial originally seeking new drugs targeted at tuberculosis noted that the drug isoniazid made patients “inappropriately happy” after a dose was administered.

A different drug, imipramine, meant to treat schizophrenia, was also found to reduce depressive symptoms.

Investigation into these drugs quickly showed that they act on biogenic amine transport and production in different ways. It was found that isoniazid and a close analog, iproniazid, work to inhibit the enzymes that breakdown monoamines, like serotonin.

Later on, it was discovered that imipramine works to block monoamine transport, which keeps serotonin outside of somatic cells longer (which is the primary way that the antidepressant SSRI drugs work). A theory started to take form – that serotonin influenced the pathological experience of depression, because limiting its degradation in the cell (iproniazid) or preventing its uptake by a neuron (imipramine) may alleviate depression.

At roughly the same time in 1955, reserpine, an alkaloid derived from Indian snakeroot (which Gandhi apparently used as a tranquilizer) seemed to increase depressive symptoms in clinical observations of just one trial. Reserpine was known to deplete biogenic amines, such as norepinephrine, serotonin, and dopamine, by blocking their transport out of the cell, in effect working in the perceived “opposite direction” of imipramine.

Looking at this in the 1950s, it would appear that if depleting biogenic amines increased depressive symptoms, and increasing biogenic amine transport via enzyme inhibition or transport reduced depressive symptoms, then a likely theory that structures these observations is that the biogenic amines themselves were responsible for depressive symptoms. By 1967, these three observations became reified in the mental health literature with the release of The Biochemistry of Affective Disorders. However, this theory, and in some cases, even the observations, were ultimately wrong, as we’ll see.

Being wrong didn’t stop what I will call the low serotonin theory of depression from fueling the next generation of innovation in anti-depression drugs, starting with fluoxetine (Prozac) and its clinical trials. The medical trade and pharmaceutical-industrial complex by now had some key factors in place in order to convince the FDA that depression is an illness, and that it is treatable. First, they reified depression as an object, with an objective criteria with which to identify and diagnose it, so that 90% of practicing physicians could identify the illness reliably. This ensured that the FDA would unequivocally recognize an indication for the drugs put forth for its review, meaning that it recognized as real, the disease of depression. This was then enshrined in the DSM, and developed further with psychiatric measurement instruments such as the Hamilton Depression Scale, which used 29 standardized questions and a computation to score the severity of your depression.

Second, they had a physiological-looking science-y theory of how it worked in the brain – the low serotonin hypothesis, which was buttressed primarily by early findings in the 1950s. What remained unsaid is that these studies were either poorly replicated, or didn’t replicate at all when run through trials decades later. Now, in order to fuel the production of first generation of SSRI drugs, we had to bring the two together – we had to see if we could drive changes in the “objective” measurement scores of depression by applying a drug to an existing population of the identified depressed. That’s precisely what the first few clinical trials used to garner FDA approval did.

If you examine their abstracts, a common theme emerges – all of the effects are framed specifically as statistical reductions in depression based on scoring with an instrument such as HAM-D. None incorporate patient experience, context, or meanings. We’ve already seen how this objectification approach maligns patient experience and meaning, but nonetheless, the score it produces is treated as an objective measurement of whether someone is depressed or not. This treatment is a decision, and the decision is made by authorities with power, over subjects to that power. This is exactly the type of thing I’m interested in calling out – it serves the power structures of medicine to heroically simplify, to build a utopian vision of how depression can be measured, much more so than the people who are subjects of this measurement, because it allows them to deploy power through institutions easily, rationally, and coherently, devaluing the individual experience of depression and preserving the authority to do so.

With the HAM-D in play, the reductions in HAM-D observed, and a theory to go along with it, the trade organizations and pharmaceutical companies were allowed to submit the drugs to the FDA for approval. The FDA recognized that the drugs served an indication, which means they are meant to treat a recognized condition, and they demonstrated effectiveness against that condition (which is another can of worms we won’t open here). Furthermore, since anyone can suffer from depression, and not just major depressive disorder, probabilities could be generated for transitioning from one score to another, putting the risk factor back into play. If you could show that a drug also plausibly reduces a risk factor that ladders up to a particular score, that multiplies the number of surrogate endpoints your drug can target, and now we are completely off the map of lived experience and physiology if you ask me. We might as well be talking about magic, flying pigs, Bigfoot, and other seemingly anecdotal and unscientific things.

The story didn’t quite end there, however. Support for the low serotonin hypothesis has been waning almost since its inception. The original reserpine studies mentioned above were replicated and re-interpreted under randomized controlled conditions, and it was found that reserpine actually did the exact opposite of what it was reported to do in 1955. The clinical criteria for evaluating whether it worked or not suffered from the same problems we discussed earlier in terms of classification and standardization. Reinterpreting the criteria under HAM-D showed no effect. Under these more standardized conditions, patients reported that they actually felt better as a result of taking reserpine. So, depleting biogenic amines in the postsynaptic neuron actually made people feel better, which also goes against the major claims (not even evidence, really) made for the efficacy of SSRIs.

Furthermore, in 1987, a clinical trial of a related compound, tianeptine, showed that enhancing serotonin uptake into the cell also alleviated symptoms of major depressive disorder, in some cases better than SSRIs such as fluoxetine. In the course of just a few years, two of the major pillars that built the low serotonin hypothesis were called into question, but the theory nevertheless persists, because it the theory was never meant to explain the experience of depression and ways to manage it. My claim, from the outset, is that the theory reified a set of medical objects that could be used to create a disease, structure the market around it, design surrogate endpoints for magic bullets to address, and then convince consumers that this particular magic bullet was responsible for solving their problems. The perceived insanity of the functioning of this market makes sense, when you stop trying to examine it logically from the patient’s perspective, and examine its logic through the lens of capital and the authorities that control it (doctors).

To recount, we’ve seen how, in order to know depression, we have to go back in and through time to see how doctors know mental health. They derived knowledge of it via classification and scientific observation, and clinical empiricism coupled with psychoanalysis and patient context. The utopian drive for objectivity ultimately won out, while the empirical approach lingered on and was used to define everyone as already ill, suffering from a nonzero amount of mental illness pursuant to the objective criterion of the DSM, and risk factors associated with each of its diagnoses. The definition of the entire population as ill or potentially ill, enabled the construction of endpoints necessary to medicalize mental disorder by the pharmaceutical-industrial complex, working in concert with the labor and complicity of the trade organizations (American Psychiatric Association and primary care physicians). This is a predictable process that, by virtue of not seeking validity, can produce knowledge that is at odds with physiology and biology, but yet can justify itself with statistics.

The epistemology of depression can be, and is, corrupted systemically. Even if every actor in the system takes the “right” action, because the system is pointed at the wrong objects, it still misfires. No malicious actors necessary, just incentives.

So what does a more effective explanatory theory of depression look like? On what side does the evidence lie? What possibilities do we as average people have to increase our ability to cope under mental stress? We’ll cover that in the remaining two parts of the series.

  1. Dumit, J. Drugs for Life: How Pharmaceutical Companies Define Our Health. Duke University Press, 2012.
  2. Pills for Mental Illness?, TIME Magazine, November 8, 1954
  3. Foucault, M. Madness and Civilization. New York: Random House. 1965.
  4. Greenberg, G. Manufacturing Depression. New York: Simon & Schuster, 2014.
  5. Chouinard, G. “A Double-Blind Controlled Clinical Trial of Fluoxetine and Amitriptyline in the Treatment of Outpatients with Major Depressive Disorder.” The Journal of Clinical Psychiatry , vol. 46, no. 3, 2, Mar. 1984, p. 32–37.
  6. Fabre, L. F., & Putman, H. P. (1987). A fixed-dose clinical trial of fluoxetine in outpatients with major depression. The Journal of Clinical Psychiatry, 48(10), 406-408.
  7. Cohn, JB. “A comparison of fluoxetine, imipramine, and placebo in patients with major depressive disorder” The Journal of Clinical Psychiatry 46(3 Pt 2):26-31 · April 1985
  8. Spitzer RL, Endicott J, Robins E. Research Diagnostic CriteriaRationale and Reliability. Arch Gen Psychiatry.1978;35(6):773–782.
  9. Endicott J, Spitzer RL. A Diagnostic InterviewThe Schedule for Affective Disorders and Schizophrenia. Arch Gen Psychiatry. 1978;35(7):837–844.
  10. Born, G.V.R., Gillson, R.E., 1959. Studies on the uptake of 5-hydroxytryptamine by blood platelets. J. Physiol. 146, 472–491.
  11. Everett, G.M., Toman, J.E.P., 1959. Mode of action of Rauwolfia alkaloids and motor activity. Biol. Psychiatry 1.
  12. Jacobsen, E., 1964. The theoretical basis of the chemotherapy of depression. In: Davies, E.B. (Ed.), Depression: Proceedings of the Symposium held at Cambridge, 22 to 26 September 1959. Cambridge University Press, New York, NY, pp. 208–214.
  13. Gram, L.F. Psychopharmacology (1986) 90: 131.
  14. Mennini, T., Mocaer, E. & Garattini, S. Naunyn-Schmiedeberg’s Arch Pharmacol (1987) 336: 478.
  15. Baumeister, A.A., Hawkins, M.F., Uzelac, S.M., 2003. The myth of reserpine-induced depression: role in the historical development of the monoamine hypothesis. J. Hist. Neurosci. 12, 207–220.
  16. Davies, D.L., Shepherd, M., 1955. Reserpine in the treatment of anxious and depressed patients. Lancet 266, 117–120.
  17. Kraepelin, E. Lectures on Clinical Psychiatry. New York: Hafner, 1968.
  18. Scull, Andrew; Schulkin, Jay (January 2009). “Psychobiology, Psychiatry, and Psychoanalysis: The Intersecting Careers of Adolf Meyer, Phyllis Greenacre, and Curt Richter”. US National Library of Medicine. 53: 5–36.
  19. Meyer, Adolf (1928). “Thirty-Five Years of Psychiatry in the United States and Our Present Outlook”. American Journal of Psychiatry
  20. Muijen, M. , Roy, D. , Silverstone, T. , Mehmet, A. and Christie, M. (1988), A comparative clinical trial of fluoxetine, mianserin and placebo in depressed outpatients. Acta Psychiatrica Scandinavica, 78: 384-390.
  21. Menninger, K. The Vital Balance. New York: Viking, 1963.
  22. Menninger, W. “Psychiatric Experience in the War, 1941-1946.” American Journal of Psychiatry, 103, no.5 (March 1947). 577-86.